In the traditional video surveillance system, the video image captured by each camera reflects only a partial view of a scene, resulting in the absence of correlation between images captured from different cameras. The movements of a moving target from a field of view of one camera to that of another camera often cause visual confusion of monitoring personnel. In particular, when the fields of view of different cameras overlap, the target will appear in multiple video images so that it is difficult for relevant personnel to track the continuous movement of the target. Since the surveillance device being widely used, the coverage of the video surveillance being expanded and the network technology being rapidly developed, there is more and more desired to have a security system capable of centralized processing, viewing and analyzing the video images from multiple surveillance cameras.